Neutron induced light-ion production from iron and bismuth at 175 MeV

(1)

Neutron induced light-ion

production from iron and

bismuth at 175 MeV

Riccardo Bevilacqua

Department of Physics and Astronomy Uppsala University

A thesis submitted for the degree of Licentiate of Philosophy

(2)

(3)

Abstract

Light-ions (protons, deuterons, tritons, 3He and α particles) production in the interaction of 175 MeV neutrons with iron and bismuth has been mea-sured using the Medley setup at the The Svedberg Laboratory (TSL) in Uppsala. These measurements have been conducted in the frame of an in-ternational collaboration whose aim is to provide the scientific community with new nuclear data of interest for the development of Accelerator Driven Systems, in the range of 20 to 200 MeV. In this Licentiate Thesis I will present the background for the present experiment, the choice of the mea-sured materials (iron and bismuth) and of the energy range. I will then give a short theoretical description of the involved nuclear reactions and of the model used to compare the experimental results. A description of the neutron facility at TSL and of Medley setup will follow. Monte Carlo simu-lations of the experimental setup have been performed and some results are here reported and discussed. I will present data reduction procedure and finally I will report preliminary double differential cross sections for produc-tion of hydrogen isotopes from iron and bismuth at several emission angles. Experimental data will be compared with model calculations with TALYS-1.0; these show better agreement for the production of protons, while seems to overestimate the experimental production of deuterons and tritons.

(4)

(5)

List of Papers

Paper I

R. Bevilacqua, S. Pomp, V. Simutkin, U. Tippawan, P. Andersson, J. Blomgren, M. ¨Osterlund, M. Hayashi, S. Hirayama, Y. Naito, Y. Watanabe, M. Tesinsky, F.-R. LeColley, N. Marie, A. Hjalmarsson, A. Prokofiev and A. Kolozhvari. Neutron induced light-ion production from iron and bismuth at 175 MeV (2009) Submitted to Radiation Measurements.

Paper II

R. Bevilacqua, S. Pomp, V. Simutkin, U. Tippawan, P. Andersson, J. Blomgren, M. ¨Osterlund, M. Hayashi, S. Hirayama, Y. Naito, Y. Watanabe, M. Tesinsky, F.-R. LeColley, N. Marie, A. Hjalmarsson, A. Prokofiev and A. Kolozhvari. Neutron induced light-ion production from iron and bismuth at 175 MeV (2009) Submitted for publication in Proceedings of the Second International Workshop on Compound Nuclear Reactions and Related Topics (CNR*09),

(6)

(7)

To the girl I met in the laundry, that night.

(8)

(9)

Introduction

To the present, all the evidence is in favor of the neutron. (James Chadwick, 1932)

1.1 History of the neutron

Since its discovery by James Chadwick in February 1932 [19], the neutron has changed the world of nuclear physics and eventually the course of human history. Not even a month passed by when in March 1932 at the Cavendish Laboratory in Cambridge, using for the first time neutrons as projectiles, it was observed that a neutron colliding with a nucleus of nitrogen could disintegrate it; two years later, in March 1934, Enrico Fermi earned his Nobel Prize discovering the possibility to induce artificial radioactivity with neutrons. Appointed Professor of Physics at Columbia University, New York, in 1939, Fermi continued his studies on neutrons; thanks to the discovery of fission, by Hahn and Strassmann [28], Meitner and Frish [42] in 1939, Fermi obtained the first controlled nuclear chain reaction, on a squash court situated beneath Chicago’s stadium, in De-cember 1942 [27]. The Manhattan Project boosted the investigation of neutron induced reactions. At the end of the Second World War, energy production became a second important driving force of neutron research, along with military applications. Others fields of study on neutron induced reactions include radiation treatment of cancer and neutron-induced nucleosynthesis in astrophysics.

The tragedies of Hiroshima and Nagasaki, along with the threat of a global nuclear war, have impressed a sign of fear in human beings [40]. The accident at the Chernobyl power plant in April 1986, has impressed in the public opinion an opposition even

(14)

to civil applications of nuclear physics [55]. In recent years the issue of treatment and disposal of spent nuclear fuel assumed crucial importance, raised interest in the political debate and concern in the public opinion; today in the European Union nuclear waste management is the main concern of the citizens opposing nuclear energy [25].

In this Licentiate Thesis I will present preliminary experimental results of inter-est for transmutation techniques in Accelerator-Driven Systems (ADS); I hope that these results will help to build safer nuclear power plants for energy production and to incinerate the radio-toxic long-lived isotopes produced in reactors of today.

1.2 ADS and transmutation

In a commercial nuclear power reactor, energy is obtained from the fission of uranium and heavier elements. The result of the fission is the creation of a large amount of neutron-rich elements; these fission products are radioactive, though most of them with short half-lives. 90Sr and 137Cs have half-lives of order of 30 years and represent the longest lived radioactive elements among the fission products in a nuclear reactor. Thus after 300 years their residual radioactivity is not a significant risk for human beings and the environment [14].

However in a nuclear reactor, a parallel process occur along with fission: elements heavier than uranium are created by neutron capture and subsequent beta decay. These elements are called trans-uranics (TRU) and include mostly plutonium and minor ac-tinides (MA) as neptunium, americium, curium, californium. The TRU elements are mostly long-lived α emitters, thus they are radio-toxic in case of human intake. Two possible solutions have been taken in account to deal with TRU elements: one possibil-ity is to dispose them in geological sites, for long periods of time (105 years), until their activity will reach the level of natural uranium; a second option is to convert the TRU elements to short-lived elements via nuclear reactions [14]. Most of the nuclear powered countries are preparing themselves for geological disposal of long lived fission products and TRU elements, including Sweden. However the geological disposal strategy rise concern in the public opinion. An overview on the Swedish geological repository and a discussion over safety issues is given in reference [26].

Transmutation is a more attractive strategy, since it allows to incinerate long-lived TRU via neutron induced reactions, however the necessary technology is not available

(15)

1.2 ADS and transmutation

yet at wide scale industrial level. A fraction of TRU elements can be treated with ther-mal neutrons, which makes transmutation in present-day reactors possible. However for a significant reduction of the long-term radio-toxicity is needed fission induced by fast neutrons. Critical fast reactors are also not suitable for incineration of all the TRU elements, in particular americium: since they produce a fraction of delayed neutrons much smaller than the one produced by 235U, they could be very difficult to control if loaded with large amounts of TRUs. A technology involving a subcritical reactor has the advantage not to require delayed neutrons for its control, however requires an external source of neutrons [11].

1.2.1 Accelerator-Driven Systems

Carlo Rubbia and his group at CERN proposed in 1993 the concept of Energy Am-plifier (EA) [18, 54]. They described this device as a high energy hadron accelerator coupled to a calorimeter, able to produce energy with a very small production of mi-nor actinides and long-lived fission products (LLFP). Two years before, in 1991, at Los Alamos National Laboratory, Bowman proposed a transmutation facility using thermal neutrons, called The Accelerator Transmutation of Waste (ATW) [17].

Accelerator Driven Systems (ADS) are a direct evolution of these two concepts. The present leading technology in ADS consists in a subcritical core coupled to a proton accelerator and a spallation target. The most promising option for the proton accelerator is to use a superconducting proton linear accelerator (Linac), operated in continuous wave (CW) mode, with energy up to 1.5 GeV and beam current of 20 mA. Lead and lead-bismuth are considered the two most promising spallation targets. As coolant are presently under investigation two options: a lead-bismuth eutectic1 or sodium.

1.2.2 Nuclear data for ADS

Transmutation techniques in ADS involve high-energy neutrons, created in the proton induced spallation of a heavy target nucleus. The existing nuclear data libraries were originally motivated by data needs in power fission reactors or fission weapons (below 8 MeV), or in fusion research or thermonuclear weapons (14 MeV); they go up to about

1_{an eutectic is a mixture of two elements at such proportions that the melting point is a local}

(16)

20 MeV, which covers all available energies for reactors of today [11]. With a spallator coupled to a core, neutrons with energies up to 1-2 GeV will be present. Although a large majority of the neutrons will be below 20 MeV, the relatively small fraction at higher energies still has to be characterized [13].

Above 200 MeV, the intranuclear cascade model work reasonably well and can be used to estimate accurately the needed cross-sections [16], however at lower energies, the incident neutron has a de Broglie wavelength corresponding to the radius of the nucleus. This implies that the entire structural properties of the nucleus participate in the nuclear reaction, making it more complex to model. This makes the 20-200 MeV region most important for new experimental cross section data [13].

These experimental data presented in this work will provide also benchmark points for state of the art theoretical models, in order to produce reliable evaluated data, to verify new phenomenological optical model potentials and to ensure a good link between low and high energy processes [36].

1.2.3 Other applications of interest

Several applications involve neutrons with energies above 20 MeV; along with energy production and transmutation of spent nuclear fuel, these include personal dosimetry in aircraft and spacecraft, radiation treatment of cancer, single-event effects in electronics. Cosmic rays reaching Earth’s atmosphere produce cascades of secondary particles in the interaction with atomic nuclei in air. Energy spectrum and intensity of the produced secondary particles depend on altitude [15], location in the geomagnetic field [34] and solar activity [43]. At commercial aviation altitudes, neutrons contribute to half of the dose received by crew and passengers of aircrafts. Dose for passenger is below the limits accepted for the public, however annual dose received by the crews of commercial aircrafts makes the latter an occupationally exposed group [46].

High energy neutrons contribute also to the dose to astronauts, residing in the International Space Station or traveling in spacecrafts, like the Space Shuttle. High energy protons, in the interaction with the material of the spacecraft or with the body itself of the astronauts, produce high energy neutrons [4]. These studies have also particular relevance for future long term space missions, including a permanent lunar base or a possible travel to Mars [23].

(17)

1.3 Choice of target materials

Fast neutron therapy is an application of neutron physics to treatment of specific forms of cancer. Fast neutrons represent an alternative to conventional radiation ther-apy with photons. Neutron are specifically able to kill cells in hypoxia and have a high linear energy transfer (LET) [57].

Single-event effects are for example radiation induced soft errors in electronic mem-ories and logic circuits. Neutron induced reactions in silicon are a major challenge for the design of high-performance microprocessors [33]. There is large economical interest in this field of research, both for terrestrial, avionic and space applications.

1.3 Choice of target materials

A large set of measurements of neutron induced light ion production cross sections at 96 MeV has been recently completed and published. These measurements were performed at the The Svedberg Laboratory (TSL), Uppsala (Sweden), using the Medley setup, that will be described in detail in Chapter 3 of this thesis. The published data include measurements on carbon[63], oxygen [61], silicon [62], iron, lead and uranium [9].

In this thesis I will present preliminary experimental results for hydrogen isotopes production at 175 MeV from iron and bismuth. These measurements were also per-formed with Medley at the quasi-monoenergetic neutron beam line at TSL.

1.3.1 Iron

Iron is an important construction material in nuclear reactors. Neutrons produced in an ADS spallation target can produce subsequently a large quantity of light-ions in the interaction with iron. Protons and α particles can cause displacements and transmutation damages in the construction materials of the reactor, in particular in the window separating the accelerator vacuum from the spallation target. These effects, coupled with mechanical stress and chemically driven processes, lower the mechanical properties of the materials used in the reactor and need to be accounted for safety issues [44]. Tritium production is also important for radioprotection issues: tritium is a volatile radioactive isotope, beta emitter, with half life of twelve years.

Iron is also an interesting element for physical considerations. Its most abundant isotope,56Fe, is the third most tightly bound nuclide, with a binding energy per nucleon of 8790.36 (±0.03) keV/A. 56Fe is an even-even nucleus, i.e. has an even number

(18)

of protons (26) and an even number of neutrons (30). Experimentally we observe that even-even nuclei are more stable than nuclei with odd numbers of protons (and neutrons).

In our experiment we have used a reaction target of natural iron; as we said, the most abundant isotope is 56Fe (91.72%), followed by 54Fe (5.8%) and 57Fe (2.2%).

1.3.2 Bismuth

Nuclei with even numbers of protons and neutrons are experimentally more stable than those with odd numbers. However some even numbers of protons and neutrons show a special behave in terms of stability; these are called magic numbers: 2, 8, 20, 28, 50, 82, 126. Nuclei with both proton and neutron magic numbers seem to be particularly favored in terms of nuclear stability, and are called double magic, for example α particles. The evidence of a special behave of nuclei with magic numbers suggests closed shell configurations (on the model of shells in the atomic structure) and supports the shell model of the nucleus.

An interesting example is calcium; this element has two double magic isotopes

40

20Ca20 and 4820Ca28, and they both have higher binding energy than the one predicted

with the Weizs¨acker formula, based on the liquid drop model. Another interesting double magic nucleus is208₈₂ Pb126. This isotope of lead is the end product of the thorium

radioactive series, and it is a particularly stable and abundant isotope. Also the end product of the neptunium, of the uranium and of the actinium series are magic isotopes, even though not double; they are respectively209₈₃ Bi126,20682 Pb124,20782 Pb125.

Natural lead is far from being a monoisotopic element: it is composed by 52.4% of 208_{Pb, 24.1% of} 206_{Pb, 22.1% of} 207_{Pb and 1.4% of} 207_{Pb. It is possible to obtain}

lead with higher concentration of a desired isotope, and with low impurities, however some companies contacted to have informations about a possible 208Pb target could offer only208Pb in form of a powder and only in milligram quantities1. Lead is also an highly toxic material [66].

We have decided though to run our experiment with a target of209Bi. This element is naturally monoisotopic, less toxic than other heavy metals, has magic number of neutrons (126) and differ just by one proton from 208Pb.

1_{Kristin Cooley, Engineering Sales Director European Region of American Elements, private}

(19)

1.3 Choice of target materials

Bismuth has been considered as spallation target in several ADS designs, eventually coupled to lead. In these accelerator designs, neutron induced light ion production from bismuth (and from lead) has the same importance than the one described for iron.

In figure 1.1 are compared a model calculation for the expected double differential cross section for light ions production from209Bi and from208Pb, induced by 175 MeV neutrons. The calculations are performed with the code TALYS-1.0; this code will be introduced in the next chapter of this thesis. We can observe that 209Bi has higher probability to produce protons with higher energies; we expected this effect, since

209_{Bi has a proton out of the magic 82 closed shell. We observe also a slightly higher}

probability in 209Bi to produce deuterons, tritons, 3He (not shown in the picture) and α particles. However the two elements show no, or little difference, at lower energies, where compound reactions occurs.

Figure 1.1: Comparison between 209Bi and 208Pb - Double differential cross section for production of protons, deuterons, tritons and α particles, from209_{Bi (red) and}208_Pb

(black), at 20◦ (solid line), 40◦ (dashed line), 60◦ (short-dashed line), 80◦ (dotted line), induced by 175 MeV incident neutrons. Model calculations with TALYS-1.0.

(20)

1.3.3 Carbon, Silicon, Oxygen, Uranium

At the The Svedberg Laboratory, we have performed several other experimental cam-paigns with Medley setup at 175 MeV. In the last two years, we have measured light-ion production induced by neutrons on carbon [29, 30], silicon, oxygen and uranium. Even though the results of these experiments will not be presented in this Licentiate thesis, they are worth a mention.

Carbon and oxygen are important elements for medical applications and for radia-tion protecradia-tion dosimetry, since they are present in human body. Oxygen in particular represent 65% in weight of human tissue. Data on neutron induced light ion production from these elements are of great interest in calculation of dose distribution in human tissue for radiation therapy with fast neutrons, as well as for dosimetry of high-energy neutrons in the upper atmosphere. 16₈ O8 is also a double magic nucleus.

Silicon data are important for detailed soft-error simulations in electronic devices, both for terrestrial application and for aircrafts, spacecrafts and satellites. The most abundant isotope of silicon is28₁₄Si14 (92%), an even-even nucleus.

Uranium represents one of the most important materials for physics of nuclear reactors. Several measurement exists of neutron induced cross sections on isotopes of uranium at energies below 100 MeV. However data are missing in the region between 100 MeV and 200 MeV, that is now of specific interest for Accelerator Drive Systems. We have measured light ion production at 175 MeV on natural uranium.

Carbon and silicon data are currently under analysis at Kyushu University (Japan). Carbon preliminary results have been presented in some conferences and published in the proceedings of those [29, 30]. Oxygen data are currently under analysis at Chiang May University (Thailand), Uranium data are under analysis at Universit´e de Caen (France).

We have performed the first experimental measurements for all these elements (C, O, Si, Fe, Bi, U) in the range of energy between 100 MeV and 200 MeV. In general, all these measurements are of great importance for benchmarking of existent nuclear models and for the specific applications previously described.

(21)

2

Elements of theory

He who loves practice without theory is like the sailor who boards ship without a rudder and compass and never knows where he may cast (Leonardo da Vinci, 1452 - 1519)

Some elements of theory of nuclear reactions and a description of the code used for model calculations are given in this chapter. The main references for section 2.1 are [56, 58, 67], while section 2.2 is referring to [37, 38].

2.1 Nuclear reactions

A nuclear reaction is a series of processes initially induced by particle collisions; the effect of this processes is either to change the intrinsic states of nuclear particles or cause their transmutations. Experimentally a nuclear reaction is induced by irradiating a target by a particle beam.

The most abundant type of nuclear reactions are processes involving a two particle collision that results in the formation of two particles; this can be written as A(a,b)B, where a is the incoming particle, A is a nucleus in the target, B is the produced nucleus and b the produced particle. The case A(a,a)A is called elastic scattering and does not involve any change in the intrinsic states of the colliding particles. All other possible reactions are inelastic processes; in this case we will observe a different final state of the involved particles and even production of particles not present in the initial state. The result of a nuclear reaction depends on the energy of the reaction itself.

In the experimental work described in this thesis, nuclear reactions are of the form A(n,xl )B, where the target nucleus A is either Fe or Bi, the incoming particle n is

(22)

a neutron, l is a light ion (isotope of hydrogen or of helium) measured by Medley; x represents one or more other particles that can be produced in the reaction along with the detected light ion l. Medley does not allow coincidence identification of more than one charged particle at time emitted in the same reaction. In these nuclear reactions it is also possible that one or more neutrons are emitted along with the mentioned light-ions, however they cannot be detected by our experimental setup. Since the scope of our experiment is to determine the double differential cross section for light-ion production, we will not further discuss the production of neutrons, nor the one of charged particles with Z>2.

To describe the processes involved in a nuclear reaction, it is convenient to consider three separate energy region in the outgoing particle spectrum: a low-energy part, an intermediate continuum and an high-energy end. This subdivision is schematically presented in Figure 2.1. The low-energy emitted particles are the result of evaporation from a compound system that has reached statistical equilibrium (C in Fig. 2.1). This region is associated to a long reaction time and to several intranuclear collisions. The particles emitted with highest energy leave the system in discrete excited states, and are the result of direct reactions (D in Fig. 2.1). These have short reaction times and involve one or two intranuclear collisions. In the between there is a broad transition region, where pre-equilibrium processes occur and where few intranuclear collisions are involved (P in Fig. 2.1).

2.1.1 Direct reactions

In a direct reaction a projectile particle interacts with just one nucleon of the target nucleus, without the formation of an intermediate compound system. In this description we will consider a nuclear shell model. A direct reaction process has highest probability to take place on a surface region of the target nucleus, without exciting the internal degrees of freedom in the rest of the nucleus. The residual particle escapes the nucleus in a time of the order of 10−22s, comparable to the time it takes a target nucleon to complete one orbit. A direct reaction involving only one intranuclear collision favors the transfer of only relatively small amounts of energy, hence it populates the ground and low-excited states of the residual nucleus. The angular distribution of the direct reaction products tend to have a forward-peaked structure. This effect increases with

(23)

2.1 Nuclear reactions

Figure 2.1: Nuclear reactions - Schematic drawing of an outgoing particle spectrum. The energy regions to which direct (D), pre-equilibrium (P) and compound (C) mechanisms contribute are indicated. The dashed curve distinguishes the compound contribution from the rest in the transitional energy region. [37]

higher kinetic energies of the projectile particle, since increases the amount of angular momentum available.

2.1.2 Compound nuclear reactions

The predominant mechanism in the production of low-energy particles is the evapora-tion from a compound system that has reached statistical equilibrium. This compound system is formed as an intermediate state; when the projectile particle and the nucleons of the target nucleus undergo several interactions they form a single entity, a compound nucleus. The time scale of this process is longer than the time scale of direct reactions, and it is of the order of 10−14s. Memory of the incident particle is not retained and the evolution of the system is determined by the amount of excitation energy available in the system. As a consequence there is no angular dependence from the incoming particle in the emitted low energy region.

(24)

2.1.3 Pre-equilibrium processes

Direct reaction and compound-nucleus are the the limiting cases, however it is not possible to define a separation line between these two processes. Nuclear reactions of intermediate nature occur. A compound system formed by an incident particle and the target nucleus may undergo disintegration before equilibrium is reached; in this case we have a pre-compound emission. Another possible mechanism is that the incident particle undergoes two or more collisions, rather than one with the target nucleons; in this case we will not have have a direct reaction, but a composite indirect reaction. Angular correlation between projectile and emitted particle is present in both cases, but stronger in the latter. Pre-equilibrium processes cover a wide range of energies in the reaction cross sections, for incident energies between 10 MeV and 200 MeV. 2.1.3.1 Multiple pre-equilibrium emission

At the energies of interest for the experimental work here described, it is necessary to consider an additional reaction process. After the first binary interaction, the residual nucleus may have enough excitation energy to allow further decay by fission or particle emission. At high incident energies this secondary decay may occur before the residual nucleus could reach the equilibrium, thus it is a pre-equilibrium process, and we identify it as multiple pre-equilibrium emission.

2.2 TALYS-1.0

TALYS is a software developed by the Nuclear Research and Consultancy Group (NRG) in Petten and the Commisariat à l’ Énergie Atomique in Bruyères-les-Châtel. The main reference for this code can be found in [38]. The purpose of TALYS is to provide state-of-the-art simulations of nuclear reactions involving neutrons, γ rays, and light ions with Z≤2 (protons, deuterons, tritons,3_{He and α particles). The code supports a}

wide energy range, between 1 keV and 200 MeV, and target masses down to A = 12. In the intentions of the authors, TALYS should enable to evaluate all nuclear reactions beyond resonance range. The code is a free software, available on-line.

TALYS includes numerous nuclear models; the user is allowed to choose among them and to adjust several parameters. However the authors indicate that TALYS can be used with default parameter for the energies, the projectile and the targets involved

(25)

2.2 TALYS-1.0

in our experiment. Hence in the simulations that we have performed with TALYS, we used the default values (blind calculations). The only informations that we provided to the code are: the projectile type, the projectile energy, the target material and the target isotopic composition. However the code allows up to 200 different keywords to be specified in the input file. The default output file include some standard cross section calculations, but it is possible to specify the desired output informations.

The TALYS Manual [37], available on-line, offers complete information on the code and the implemented nuclear models.

2.2.1 Examples of model calculations

Here are presented some examples of model calculations performed with TALYS-1.0. In figure 2.2 are plotted the neutron induced proton production cross sections calculated by TALYS, on a target of 56Fe (a) and of 209Bi (b), with a projectile energy of 175 MeV for an emission angle of 20◦ in the centre of mass system.

TALYS provides separately the contributions of the different nuclear reactions mech-anism; in figure 2.2 are plotted separately contributions from compound reactions, pre-equilibrium emission, multiple pre-equilibrium emission and direct reactions. The output file contains also the total double differential cross section for each requested emission angle in the laboratory system.

2.2.2 Laboratory system vs. center of mass system

Double differential cross sections (DDX) are calculated by default in the center of mass (CM) system. It is possible to use TALYS to calculate DDX in the laboratory (LAB) system, that is the system in which the experimental cross sections are obtained. However computational times differ by three orders of magnitude: to calculate the DDX for production of protons from iron at 20◦ in the CM system, TALYS employed 120 seconds; on the same machine, a calculation of the DDX for proton production from iron at 20◦ in the LAB system lasted 20 hours (7.2×104 seconds).

This computational time difference is due to the fact that to one emission angle in the CM system correspond an emission angular spectrum in the LAB system and that to one emission angle in the LAB system correspond an emission angular spectrum in the CM system; since the TALYS code computes the DDX in the CM system, to obtain results expressed for one angle in the LAB system is necessary to calculate the DDX

(26)

Figure 2.2: TALYS example. - Double differential cross section for56Fe(n,xp) (a) and of209_{Bi(n,xp) (b) reactions, at 20}◦_{in the centre of mass system, and with incident energy}

of 175 MeV. The solid red line represent the sum of the following contributions: compound reactions (dashed light blue), pre-equilibrium (dashed blue) and multiple pre-equilibrium emission (dashed pink), direct reactions (dashed green)

in all the angles in the CM system and to sum all the contributions to the requested emission angle in the LAB system.

Experimental data in the present work includes production of three hydrogen iso-topes and two helium isoiso-topes, at eight angles in the LAB system, from two reaction targets. Calculations in the CM system with TALYS will require 1600 hours to be completed. However the mass of the produced particles considered in this experiment is small compared to the mass of the residual nuclei, thus we expect the difference between CM system and LAB system to be small.

In figure 2.3 we present a comparison between DDX calculated with TALYS in the LAB system and in the CM system; we can observe that at small angles and for protons the difference is very little, while for tritons at larger angles calculations if we assume the CM system to be equivalent to the LAB system we overestimate the production cross section at higher energies. Since all the comparisons between experimental data and TALYS calculations presented in this thesis will be between CM system calculations and experimental LAB system measurement, it will be necessary to take in account this effect.

(27)

2.2 TALYS-1.0

Figure 2.3: LAB system vs. CM system - In the left panel, comparison between double differential cross sections for production of protons at 20◦ from iron in the LAB system (black line) and in the CM system (red line). In the right panel, same comparison but for production of tritons at 80◦.

(28)

(29)

3

Experimental Methods

Though this be madness, yet there is method in ’t. (William Shakespeare, Hamlet, Act 2 scene 2)

3.1 Neutron facility at the The Svedberg Laboratory

The The Svedberg Laboratory (TSL), is a university facility based in Uppsala, Swe-den. The activity of TSL is mainly devoted to proton radiation therapy for cancer. Beam-time not used for proton therapy is devoted to commercial neutron and proton irradiation projects. The facility is also available for basic academic research for project associated to Uppsala University or to EU programs.

The laboratory is named after The (Theodor) Svedberg (1884-1971) professor in physical chemistry at Uppsala University and Nobel Prize laureate in chemistry (1926). In 1945, a donation from the Gustaf Werner Corporation gave the opportunity to build a large particle accelerator, a synchro-cyclotron. This cyclotron is the same still used today for accelerating charged particles at TSL.

The first neutron facility was built at TSL in the late 1980s [21] and remained in operation until 2003. A new neutron beam facility has been then constructed in the following year and has been available from 2004 [48, 50].

3.1.1 Neutron production

At low energies it is possible to produce truly monoenergetic neutron beams. Possible reactions to obtain such a neutron beam are2H(d,n)3He and3H(d,n)4He. However only incident deuterons with energy up to 2 MeV produce monoenergetic neutron beams. For

(30)

higher energies the deuteron can break up in a proton and a neutron, even though up to 30 MeV the cross section for the reaction3H(d,n)4He is much larger than the probability for a deuteron break up, thus producing a small low-energy tail. For even higher energies the3H(d,n)4He cross section is too small to provide a total yield sufficient for most measurements [11, 12].

For energies above 20 MeV it is possible to obtain a neutron beam with with a strong dominance of neutrons with a specific narrow energy, plus a broad low-energy tail. A beam with these characteristics is called quasi-monoenergetic; three proton induced nuclear reactions are available to produce a quasi-monoenergetic neutron beam:

2_H(p,n), 6_Li(p,n),7_{Li(p,n). Most of the facilities available to the scientific community}

uses the reaction on 7Li, since6Li is a controlled isotope for safeguard issues1 and the reaction on 2H does provide a broad high-energy peak [11, 12].

3.1.2 Neutron beam line at TSL

At TSL neutrons are generated via 7Li(p,n) reaction; protons are accelerated in a cyclotron, extracted and focused in a beam of well defined energy and intensity. The proton beam is transported in the experimental hall (Blue Hall) where it is impinging on a 99.99% enriched7Li target; the peak monoenergetic neutrons are produced by the

7_Li(p,n)7_{Be reaction, while the low energy tail is produced via other channels, opened}

by the high energy of the incident protons. There are several available lithium targets, of different thicknesses: 1, 2, 4, 8.5 and 23.5 mm. In the present experiment the nominal proton energy was 179.3 (±0.8) MeV, the lithium target thickness 23.5 mm and the average energy of high-energy peak neutrons 175.0 (±2.5) MeV. The residual proton beam is deflected by a bending magnet into a beam dump; here the proton beam is integrated in a Faraday cup in order to monitor the beam current.

A neutron beam is shaped with a set of iron collimators, as described in Figures 3.1 and 3.2. A first cylindrical iron collimator, with inner diameter of 20 mm and a length of 400 mm, is shaping the beam and is determining the beam size at the interaction

1_{In a fusion nuclear weapon, energy is produced via the deuterium-tritium reaction:} 2_{H +}3_{H →} 4_{He + n + 17.6 MeV. To obtain the tritium needed for this reaction, the}6_Li(n,t)4_{He is used. Nations}

with nuclear weapon arsenals keep strategic reserves of6Li. The United States, for example, produced a total of 442.4×103 Kg of enriched lithium from 1954 to 1963 for thermonuclear weapons, tritium production, and other purposes; they suspended the production of6_{Li in 1963. Production, export and}

(31)

3.1 Neutron facility at the The Svedberg Laboratory

Figure 3.1: Collimator design (a) - Neutron beam line, collimator and Medley spec-trometer (TSL drawing, private communication).

(32)

Figure 3.2: Collimator design (b) - Detail of the collimator design. Drawing axis have different scales (TSL drawing, private communication).

(33)

Figure 3.3: Collimator data sheet - Reproduction of the technical data sheet with description of the collimator’s configuration (A. Prokofiev, private communication)

(34)

with the target materials. A second set of conical iron collimators, with upstream inner diameter of 30.09 mm, downstream inner diameter of 54.00 mm and total length of 1375 mm, is positioned afterward to shield the beam halo. A third conical iron collimator, 250 mm long, with 61.46 mm upstream inner diameter and 65.81 downstream inner diameter, is also placed in the beam line. In Figure 3.3 we reproduce the original technical data sheet from TSL with the description of the collimator’s configuration.

The proton beam line, the lithium target and the bending magnet are closed in a concrete bunker to shield the experimental area from background radiation. To shield the Medley setup from background radiation, the concrete wall downstream the neutron line has been replaced by a 100 cm thick iron wall. The shielding has been further upgraded with a second 50 cm thick iron wall. Inside the bending magnet, downstream the lithium target, lead bricks have been positioned around the neutron beam pipe to further improve the shielding.

In figure 3.4 are presented quasi-monoenergetic neutron spectra measured at the TSL neutron beam line, compared with model calculations by Pomp et al. [48]. Incident proton energy, thickness of the 7Li target, peak energy of the produced neutrons and fraction of those over the total, for each case, are reported in Table 3.1 [50]. In the same table are also reported experimental values measured in the work presented in this thesis. These measurements of the incoming neutron energy spectrum will be described in Section 5.4; time-of-flight (TOF) data are used in the off-line analysis to select peak neutrons and reject (part of) neutrons in the low-energy tail and will be discussed in Section 5.5.

Incident proton 7Li target Average energy of Neutrons in the energy (MeV) thickness (mm) peak neutrons (MeV) high-energy peak(%)

24.68± 0.04 2 21.8 ∼ 50

49.5± 0.2 4 10 39

97.9± 0.3 8.5 94.7 41

147.4± 0.6 23.5 142.7 55 (upper limit)

179.3± 0.8 23.5 174.9 39

Table 3.1: Parameters of quasi-monoenergetic neutron beams at TSL - Data for proton energies up to 147.4 MeV have been measured by Prokofiev et al. [50]. Data for Ep= 179.3 MeV have been measured with Medley setup, as part of the experimental work

(35)

Figure 3.4: Quasi-monoenergetic neutron spectra - Measured neutron spectra at 0◦ for different peak neutron energies at the TSL neutron beam line. Symbols connected by a solid line represent experimental data, while theoretical calculations are shown as dashed lines [48]

(36)

3.1.2.1 Neutron beam monitors

The relative neutron beam intensity is monitored by two separate devices, both of them utilizing neutron induced fission of 238U: a thin film break down counter (TFBC) [52] and an ionization chamber (ICM). These monitors are both positioned downstream the Medley experimental setup. However, as we will see in the next Section, Medley is an evacuated chamber and the reaction target is thin enough to not significantly affect the neutron beam. In addition, as previously mentioned, the current of protons collected at the beam dump is integrated in a Faraday cup, to provide a third (indirect) monitor of the neutron flux.

3.1.2.2 Why 175 MeV?

Why 1/137? (Wolfgang Pauli,1900-1958)

A last question needs to be answered: after defining the energy region of interest (100 to 200 MeV) and the neutron energies available at TSL (11 to 175 MeV), why did we perform our measurements at exactly 175 MeV? This energy is the maximum neutron energy available at TSL. While this does not justify the choice to use it to run our experiment, we should also note that the choice of a specific energy in the region of interest includes always a portion of arbitrariness. However there is, if not a reason, at least a justification for our choice. As we mentioned before, the main activity of TSL is devoted to proton radiation therapy for cancer; in order to easily access the beam line and to have more beam time for our experiment, we performed our measurements in beam sharing mode with the radiation therapy. This means that the proton beam is switched between the radiation therapy line and the neutron line during the operation time. This switch can be done in few seconds, however since the main user is the radiation therapy, this requires to use a specific proton energy.

(37)

3.2 Medley setup

This medley of philosophy and war. (Joseph Addison, Cato, Act 2 scene 6)

Medley1 is a spectrometer system, semi-permanently installed at the TSL neutron beam line. Medley consists of eight three-element detectors installed in a cylindrical evacuated scattering chamber, positioned symmetrically downstream the neutron beam line. The reaction chamber has an internal diameter of 800 mm and a height of 24 cm. A schematic drawing of Medley is presented in figure 3.5 (a), while figure 3.6 is an actual picture2 of the chamber, with the installed detectors.

Figure 3.5: Medley setup - (a) The reaction chamber, the arrangement of the eight detectors and of the target. (b) Construction details of each three-element telescope de-tector.

The chamber has four ports; two of them are aligned with the neutron beam line and are used for beam transport, a third one is used to evacuate the chamber with a vacuum turbo pump, and the last one is used to hold a calibration source. Inside the chamber there are eight telescope detectors; each telescope is mounted on an individual rail, aligned on a radius of the chamber. There are eight rails, separated in steps of

1

in some publications Medley was written with capital letters (MEDLEY). However the word Medley is not an acronym, thus we decided here to use a more consistent spelling.

2

An acute observer will notice the concrete wall on the left side of the image. This picture was taken before the construction of the new iron wall, however the internal configuration of the reaction chamber has not changed since then.

(38)

Figure 3.6: A picture of Medley - An actual picture of the reaction chamber and of the telescope detectors. On the upper left side is possible to see the entrance of the neutron beam.

20◦; these are fixed on a turnable plate at the bottom of the chamber. The plate can be operated externally, without breaking the vacuum, and can be turned by 360◦; an indicator and a scale allow the telescopes to be positioned at the desired angles. In principle every scattering position in the laboratory system can be measured. However there are some limitations: no telescope can occupy angles smaller than 20◦, nor larger than 160◦, to prevent the telescopes to be directly irradiated by the neutron beam.

In the standard configuration the telescope detectors are positioned at 20◦, 40◦, 60◦, 80◦, 100◦, 120◦, 140◦, 160◦. Calibration runs are also performed with different configurations. The position of the telescopes on the rails is adjustable, so that the distance between the telescopes and the reaction target can vary between 180 mm and 250 mm. The telescopes are identified by numbers, from T1 to T8; in standard configuration T1 is positioned at 20◦, and T8 at 160◦. Half of the beam time has been used in standard configuration, while in the other half of the beam time we turned the plate by 180◦, thus having T8 at 20◦.

At the centre of the chamber a reaction target is positioned, mounted on a thin aluminum frame, 200 mm wide and 140 mm high, using thin threads. The distance between the lithium target and the reaction target, in the present configuration, is 4618

(39)

3.2 Medley setup

mm; at this position the quasi-monoenergetic neutron beam has a diameter of 42.08 mm, thus allowing free passage of the beam through the target’s frame. Three frames are installed in the chamber, each of them supporting a different reaction target. The frames can be operated externally, thus allowing to switch the target at the centre of the chamber without breaking the vacuum. A fourth configuration is available, removing all the frames from the central position, and it is used to measure instrumental background or to insert a calibration source. During operation the chamber is evacuated and the internal pressure is kept below 10−5 mbar. Signal cables for each detector are brought out of the Medley chamber via connectors mounted at the bottom of the chamber itself. 3.2.1 Telescopes

Each telescope detector installed in Medley consists of two fully depleted silicon surface detectors and a CsI(Tl) scintillator; a schematic drawing of a telescope detector is pre-sented in figure 3.5 (b). This configuration has been chosen to obtain a good particle identification with low threshold and over a wide dynamic range. In the present con-figuration Medley can measure and identify protons with energies from 2 MeV up to 180 MeV; for other ions thresholds are higher, up to 9 MeV for α particles. ∆E-∆E-E technique is used, and it will be discussed in Section 5.1. The two silicon detectors in each telescope have different thicknesses, one being thinner than the other; they are identified respectively as Si1 and Si2 and are used as ∆E detectors. The scintillator is

used to fully stop the incoming charged particle and serves as E detector. 3.2.1.1 Silicon detectors

The silicon detectors are fully depleted standard silicon surface barrier detectors from ORTEC. Si1 is the thinner silicon detector in each telescope and it is always placed in

the front position, closer to the reaction target. Thickness of Si1 varies between 50 µm

and 65 µm. The second silicon detector (Si2) is 1000 µm thick. This has been a recent

upgrade of the Medley setup; in previous experiments at 175 MeV we used 500 µm to 600 µm Si2 detectors [29]. To improve identification of high energy light-ions, we

have subsequently installed new thicker Si2 detectors. However one of the new silicons

(installed in T4) has shown some malfunctioning, and has been replaced with an older 500 µm one. As we will discuss in Chapter 6, a second 1000 µm (installed in T7) showed an anomalous behavior, unfortunately discovered only in the data reduction process.

(40)

Thickness of Si1 and of Si2 for each telescope is reported in Table 3.2; in the same

table punch-through energies for protons, deuterons and tritons, calculated with the SRIM code [68, 69] are reported.

3.2.1.2 CsI(Tl) scintillator

The E detectors are 100 mm long CsI(Tl) scintillators; they have cylindrical shape, with a diameter of 50 mm, where the last 30 mm are tapered to 18 mm diameter to match the size of a 18 × 18 mm2 Hamamatsu S3204-08 [22] photodiode for the light read out. The emission spectrum of CsI(Tl) has its maximum at 550 nm at it is poorly matched to the response of photomultiplier (PM) tubes; however using photodiodes, the scintillation yield is higher than the one of any other scintillator [35]. In addition the use of a photodiode to collect the light from the CsI(Tl) allow a more compact design, suitable for the dimensions of the Medley reaction chamber.

In previous experiments at 96 MeV [9, 61, 62, 63], Medley was equipped with 50 mm long CsI(Tl). While these scintillators were enough to fully stop all the light ions produced in the interaction of 96 MeV neutrons with target materials, they were not suitable for the new energy range. Calculations made with the SRIM code shows that the new 100 mm long CsI(Tl) are sufficient to fully stop all the produced charged particles in the 175 MeV interaction.

3.2.2 Reaction targets

Three reaction targets were mounted in the scattering chamber; an iron and a bismuth target were used to measure respectivelynatFe(n,xl ) reactions and209Bi(n,xl ) reactions. As third target we used polyethylene (CH2) to measure the H(n,p) elastic reaction for

absolute cross section normalization (discussed in Section 5.6) and to calibrate the CsI(Tl) scintillators (Section 5.2).

The Fe target was 1959.6 (±0.1) mg of natural iron, 375 µm thick, with a square surface of 25 × 25 mm2. The Bi target, naturally monoisotopic, had a mass of 3130.1 (±0.2) mg, was 0.5 mm thick and had a square surface of 25 × 25 mm2_{. The CH}

2

target, 1.0 mm thick, with a mass of 461.55(±0.1) mg and a diameter of 25 mm, was the same used in previous experiments with Medley at 175 MeV [29].

The reaction targets are mounted on individual frames, and they have an inclination of 45◦ respect to the beam incident direction. As previously mentioned, the neutron

(41)

3.2 Medley setup

Protons thickness ESi1 thickness ESi2 Si1 + Si2 ∆ESi1

Si1 (µm) (MeV) Si2 (µm) (MeV) (µm) (MeV)

T1 64.9 2.394 1026 12.249 1090.9 0.442 T2 60.5 2.292 1018 12.179 1078.5 0.414 T3 63.9 2.337 1012 12.156 1075.9 0.438 T4 52.9 2.091 549 8.524 601.9 0.473 T5 50.4 1.990 1008 12.150 1058.4 0.347 T6 50.1 1.990 962 11.846 1012.1 0.351 T7 61.7 2.292 963 11.861 1024.7 0.431 T8 53.4 2.091 1016 12.228 1069.4 0.365

Deuterons thickness ESi1 thickness ESi2 Si1 + Si2 ∆ESi1

T1 64.9 3.090 1026 16.497 1090.9 0.599 T2 60.5 2.933 1018 16.433 1078.5 0.560 T3 63.9 3.056 1012 16.321 1075.9 0.592 T4 52.9 2.695 549 11.454 601.9 0.638 T5 50.4 2.591 1008 16.322 1058.4 0.470 T6 50.1 2.591 962 15.920 1012.1 0.476 T7 61.7 2.991 963 15.907 1024.7 0.585 T8 53.4 2.695 1016 16.398 1069.4 0.496

Tritons thickness ESi1 thickness ESi2 Si1 + Si2 ∆ESi1

T1 64.9 3.494 1026 19.475 1090.9 0.717 T2 60.5 3.392 1018 19.424 1078.5 0.670 T3 63.9 3.494 1012 19.386 1075.9 0.708 T4 52.9 3.093 549 13.517 601.9 0.760 T5 50.4 2.992 1008 19.329 1058.4 0.561 T6 50.1 2.992 962 18.825 1012.1 0.569 T7 61.7 3.430 963 18.793 1024.7 0.700 T8 53.4 3.093 1016 19.398 1069.4 0.593

Table 3.2: Energy deposition of hydrogen’s isotopes in Si detectors. Thickness of each ∆E Si detector is reported. ESi1 and ESi2 are punch-through energies of

pro-tons, deuterons and tripro-tons, calculated with the SRIM code [68, 69]. ∆ESi1 is the energy

(42)

beam has a diameter of 42 mm at the centre of Medley, thus all the three reaction targets used in this experiment are fully covered by the incident beam.

3.3 Readout, electronics and data acquisition system

3.3.1 Readout and electronics

We provide here a short description of the readout and electronics, that is essentially the same used in all previous experiments with Medley at TSL [9, 29, 61, 62, 63]. In figure 3.7 is reported a scheme of the electronics.

We will identify the signals as A, B, C respectively for Si1, Si2, CsI(Tl); we will also

use T to indicate a time signal and E to indicate an energy signal. Finally numbers 1 to 8 refer to Telescope 1 to Telescope 8.

Analog signals from Si1, Si2 and CsI(Tl) are pre-amplified in the experimental area

(Blue Hall) and transported to the counting room where all the electronics and the data acquisition system are placed. Signals from Si1 and Si2 are split, to be used as time

signals (TA1, TA2, ..TA8, TB1, ..TB8) and as energy signals (EA1, EA2, ..EA8, EB1, ..EB8), in order to build a trigger. Time information is not used for the scintillator, since it has too slow rise time to be useful, thus we have only energy signals (EC1, ..EC8).

3.3.1.1 Energy signals

The 24 energy signals (three per telescope), once transported to the counting room, are amplified with a Dual Amplifier by ORTEC (Dual Spec Amp 855) and fed to a 32 Channel Multievent Peak Sensing ADC by CAEN. A Master Trigger originated from the time signal will allow the gated events to be sorted and stored on a drive.

3.3.1.2 Time signals

Signals from Si1, Si2 are converted to NIM logic pulses by constant fraction

discrimina-tors (Quad CFD, TC 455 by Tennelec and 934 by ORTEC). These devices are designed to produce timing information from analog signals. The incoming analog signal is split; one of the two identical signals is then attenuated, delayed and inverted respect to the other. The two signals are then fed to a fast comparator and a timing signal is triggered at a constant fraction of the input amplitude.

(43)

3.3 Readout, electronics and data acquisition system

The 16 time signals (TA1 to TB8) are then grouped with a series of Logic FIFO (Fan In, Fan Out), to create a single Master Event. The Master Event is then fed to a coincidence unit (LeCroy 465) and is vetoed by the Computer Busy signal generated from the crate controller of the acquisition system. The logic signal is then fed to a second coincidence unit where it is vetoed by an RF-off signal (see below). The logic output is the Master Trigger used to gate the energy signals in the ADC, to start the TDC units, to trigger the Crate Controller. The 16 time signals from the CFDs are also fed to an array of TDC (LeCroy 2228).

Protons are extracted from the cyclotron in separate bunches, with length of few ns and a repetition time of ∼45 ns. This repetition time is given by the RF signal of the cyclotron and it is used as a reference signal for time-of-flight (TOF). We measure the TOF as time difference between a registered event and the next RF pulse, hence what we actually measure is a negative TOF. The RF-off signal is generated from the RF signal through a timer and a logic FIFO, to veto the acquisition of data when there are no proton bunches producing neutrons.

3.3.2 Data acquisition system

A description of the data acquisition system SVEDAQ used at TSL and a User’s Guide are available on-line [45]. SVEDAQ is based on the Multi Instance Data Acquisition System (MIDAS) software [53]. An Event Builder system, composed by a CPU, a CAMAC interface and a Data Network interface is located in a VME crate. The Event Builder reads out the ADC and the TDC, collects the events in buffers and send them to a Disk Server via the Data Network. This is a dedicated network based on a 10 Mbit/s Ethernet. The Disk Server is also located in the VME crate and it is connected to a disk drive to store the data from the experiment. The data acquisition system is controlled by a SUN workstation. Through the same workstation is possible to monitor the signals from the CAMAC system and to perform some on-line analysis.

(44)

Figure 3.7: Scheme of Medley electronics - A general scheme of the electronics used in the data acquisition system for the Medley experiment.

(45)

4

Monte Carlo calculations

The Monte Carlo method for solving transport problems emerged from work done at Los Alamos during World War II. (from MCNP User’s Manual)

The neutron beam line at TSL has been upgraded in the last years to provide a quasi monoenergetic spectrum with higher peak energy at sufficient intensity [48, 50]. Since January 2004 a maximum energy of 175 MeV is available. The neutron produc-tion lithium target has been moved closer to the experimental area, where Medley is positioned. This, together with the higher energy of produced neutrons, has signifi-cantly increased the neutron background in the experimental area, affecting the signal to noise ratio in the Medley measurements. Two iron walls, respectively 100 cm and 50 cm thick (see figure 3.1), have been installed to shield Medley from the lithium target, as described in the previous Chapter. Here I will describe further improvement of the shielding and the choice of the collimator shape, according to the results of Monte Carlo simulations of the experimental setup performed with MCNPX 2.5.

4.1 MCNPX

MCNPX is a general purpose Monte Carlo radiation transport code developed at Los Alamos National Laboratory (New Mexico, USA). MCNPX code is an extension to all particles and all energies of the MCNP (Monte Carlo N-particles) code, in which only transport of neutrons (up to 150 MeV), photons (up to 100 GeV) and electrons (up to 1 GeV) is implemented. In MCNPX is possible to choose which data libraries to use for the simulations; in the present work, the LA150 evaluated data libraries have

(46)

been used for protons and neutrons in the energy region below 150 MeV [20], while for energies above 150 MeV the Intranuclear Cascade model by Bertini has been employed [6].

In general terms a Monte Carlo method generates a set of particle tracks and follows them according to the statistical rules of the cross section database. Cross-section data are used as probabilities for interactions. Ray-tracing through geometry of the problem is used to determine the location of those interactions. Events that occur during the simulation are stored. Physical observables in the simulation can by obtained using different tallies; a tally is a function in MCNPX that asks to the code to register a physical observable (e.g. flux, energy deposition) in a specific user defined region of the geometry. An input file containing the geometry of the problem, informations on the materials, a description of the source (primary particles) is provided to the code. The input file includes also informations about tallies, energy and geometry cut off, variance reduction techniques.

In the present work, the geometry of the neutron beam line included the collimators, the shielding iron walls, the concrete bunker, the beam pipe, the Medley chamber and the eight telescope detectors in it. As primary source particles we used neutrons, generated in the position of the lithium target (not present in the simulation). The source neutron spectrum implemented in the simulation consists of 40% neutrons in the 175(±2.5) MeV peak and 60% uniformely distributed between 1 MeV and 172.5 MeV. Neutrons whose energy, after any interaction, fell below 1 MeV were terminated (energy cut off). Simulated neutrons where emitted in a solid angle of 8◦ from the lithium target position (geometry cut off).

Several tallies have been applied in different areas of the simulated geometry to record the particle flux and energy spectrum. Flagged geometry regions allowed to identify the contribution to the flux and to the spectrum of particles coming from specific areas.

4.2 Background studies

4.2.1 Extra shielding

A set of simulations has been performed to identify and eventually mitigate the major contributors to the background neutrons in the Medley chamber. We recorded neutrons

(47)

4.2 Background studies

reaching the telescope detectors: these are not in the line of view of the collimator opening, thus the neutrons should have leaked through the shielding iron walls, should have scattered on the collimator surface or should be secondary neutrons produced by interaction of primary particles. To selectively identify the background neutrons we flagged several surfaces in the simulated geometry: this procedures allows to consider only particles passing (or not passing) through specific surfaces, i.e. allows to identify the origin of the detected particles and to track their path.

The simulations have shown that 50% of the neutrons recorded in the telescope at 20◦ are primary neutrons originated from the lithium target. This fraction increases up to 60% for the telescope at 80◦ (whose surface is geometrically moore exposed) and for the telescope at 160◦ (that is closest to the neutron source). We have simulated then the insertion of an extra lead shielding to be inserted in the bending magnet, around the neutron beam pipe, downstream the bending of the proton beam pipe. This is the closest position available to shield the lithium target in the actual experimental configuration. With this extra shielding, the simulations showed a reduction of 30% of background neutrons recorded in the telescopes. In figure 4.1 is shown the ratio between neutrons recorded in the telescope at 160◦ without and with the insertion of an extra lead shielding inside the bending magnet.

This shielding solution has been implemented1 and it is currently installed at the neutron beam line at TSL.

4.2.1.1 Discarded solutions

Insertion of an extra lead shielding inside the Medley chamber has been also simulated. This solution has also been tested experimentally, inserting some lead bricks in the reaction chamber; the limited space inside Medley allows to shield only the four forward telescopes. However MCNPX simulations showed that in the forward telescopes a small reduction of high energy background neutrons was compensated by an increase of medium and low energy background neutrons in both forward and backward detectors. Hence this solution has been discarded.

A set of simulations with insertion of a paraffin wall between the two iron shielding walls separating Medley from the lithium target has been performed. Also this solution

1_{actually this solution has been supported by Monte Carlo simulations ex post its first application}

(48)

Figure 4.1: Lead shielding effect at 160◦ - Ratio between neutrons recorded in the telescope at 160◦_{before and after the insertion of an extra lead shielding inside the bending}

magnet, downstream the lithium target.

has been tested experimentally at TSL. The simulations with MCNPX did not show any significative reduction of the background neutrons. Also this solution has been discarded.

4.2.2 Collimator configuration

At TSL are available conical and cylindrical sets of modular collimators, with several opening diameters; these can be inserted in the neutron beam pipe to shape the neutron beam and to shield the beam halo. A second set of simulations has been performed to determine the best collimator configuration.

Simulations have been performed considering two sets of cylindrical collimators, with 25 mm diameter opening and with 35 mm diameter opening, and a conical colli-mator with opening diameter from 23.75 mm to 70.16 mm. The Monte Carlo calcula-tions showed a 50% reduction in the number of background neutrons recorded in the telescopes using a 25 mm diameter cylindrical collimator instead of one with a diameter of 35 mm. The use of a conical collimator showed a further reduction of background neutrons. Telescopes at 20◦ and 160◦, that are closer to the beam, recorded less than a factor 10 of background neutrons in the conical configuration compared to the 35 mm cylindrical one. Telescopes at angles between 40◦ and 140◦ showed also a reduction of

(49)

background neutrons: this simulation produced 20% of the neutrons recorded with the 35 mm cylindrical configuration at energies below 100 MeV and 10% at higher energies. A mixed configuration has been suggested to avoid any hot surface close to Medley: the beam is shaped only by a first cylindrical part, 400 mm long and with a diameter of 20 mm; a second conical collimator is positioned downstream with initial inner diameter of 30.09 mm (thus larger than the shaped beam), to shield the beam halo. The simulation results showed that this configuration is equivalent to the conical one for telescopes at angles between 40◦ and 140◦, however the two telescopes closer to the beam (at 20◦ and 160◦) recorded a significant lower number of background neutrons, over the wide energy spectrum. This mixed configuration has been selected for the actual experiment and it is the one described in the previous Chapter and in figures 3.1 and 3.2.

In figure 4.2 are presented the MCNPX calculation results, for the four collimator configurations here described, in terms of ratio between background neutrons recorded in the telescope at 80◦ with a 35 mm diameter cylindrical collimator and with, re-spectively, a 25 mm diameter cylindrical collimator, a conical collimator and a mixed cylindrical-conical collimator. In figure 4.3 the same ratio is presented but for back-ground neutrons recorded in the telescope at 160◦. It is possible to observe that the mixed configuration provide the lowest neutron background in both forward and back-ward telescopes, in particular for the telescopes closer to the beam line.

4.2.3 Proton background

Neutrons are not directly detected by the three-element telescope detectors installed in Medley, however protons and other charged particles are created in the interaction between neutrons and the materials present in the experimental area. A set of Monte Carlo simulations has been performed to investigate the major contributors to proton background in the Medley experiment and to characterize this background.

Primary neutrons where generated at the lithium target position and transported through the geometry. When a proton was generated by interaction of a primary (or secondary) neutron with collimators, with Medley chamber or with the CsI detec-tors, this was transported and its eventual energy deposition in the telescope detectors recorded.

(50)

Figure 4.2: Comparison between different collimators in telescope at 80◦- Ratio between background neutrons recorded in the telescope at 80◦with a 35 mm diameter cylin-drical collimator and with, respectively, a 25 mm diameter cylincylin-drical collimator (blue), a conical collimator (red) and a mixed cylindrical-conical collimator (green).

Figure 4.3: Comparison between different collimators in telescope at 160◦ -Same as figure 4.2 for the telescope at 160◦.

(51)

4.2.3.1 Interaction with Medley chamber

The Medley chamber has 50 mm thick aluminium walls. Background neutrons inter-acting with the chamber produce protons and other charged particles. The range of 175 MeV protons in aluminium calculated with SRIM [69] is 97.56 mm, thus protons produced in all the thickness of the chamber could reach the telescope detectors and trigger an event. Simulations showed that these protons contribute to less than 1% of the total number of protons recorded in the telescopes.

4.2.3.2 Interaction with CsI scintillators

Neutrons can interact with the CsI(Tl) scintillators and produce protons, other charged particles, γ and neutrons; in this simulations proton production has been considered. The protons produced in the scintillator will then deposit here their energy and can trigger an event reaching the silicon detectors. The average proton flux recorded in the simulated CsI detectors is presented in figure 4.4. The contribution to the proton background in telescopes, by neutron induced proton production in the CsI scintillator has been calculated to be less than 10% of the total.

Figure 4.4: Background proton production in CsI - Average proton flux induced by background neutrons recorded in simulated CsI detectors at 20◦_{, 40}◦_{, 60}◦_{, 80}◦ _(unit

(52)

4.2.3.3 Proton production in the collimator

The new collimator design, with an initial cylindrical section to shape the beam fol-lowed by a longer conical section to shield the beam halo, reduced the number of protons created in the direct interaction of beam neutrons with the collimator internal surface. However simulations performed with MCNPX showed that the contribution of back-ground protons recorded in the scintillators created by interaction of neutrons with the iron collimators is about 90% of the total. In figure 4.5 is compared the average proton flux in the telescope detector at 20◦ due to neutron induced reactions in the collimator and in the CsI scintillator.

Figure 4.5: Background proton production in the collimator - Comparison between proton background induced by neutrons interacting respectively with the collimator (red) and with the CsI scintillator (blue), as recorded in the telescope detector at 20◦ _(unit

Neutron induced light-ion production from iron and bismuth at 175 MeV